Advanced Kinetic Modeling of Waste-to-Energy Conversion from Commodity Waste Plastic Pyrolysis Using Thermogravimetric Analysis and Predictive Artificial Neural Network Models

This study explores the kinetic properties of four waste polymers, polyethylene terephthalate (PET), polystyrene (PS), polyvinyl alcohol (PVA), and polyvinyl chloride (PVC), using thermogravimetric analysis under an inert environment between 300 and 973 K at three different heating rates. Several model-free isoconversional methods, including the advanced isoconversional method, along with the model-fitting Coats−Redfern method, were applied to determine key kinetic parameters: activation energy (Eα), pre-exponential factor (Aα), and reaction mechanism (f(α)). Thermodynamic parameters were also calculated to gain insights into the degradation behavior, crucial for optimizing large-scale pyrolysis reactor designs. The average activation energies calculated using advanced isoconversional methods were 208.50, 198.14, 118.79, and 117.12 kJ mol−1 for PET, PS, PVC, and PVA, respectively. The governing reaction mechanisms that were identified based on Criado masterplots were F2 for PET, R3 for PS, F2 for PVC, and F3 for PVA, indicating a solid-state reaction mechanism for each polymer. Additionally, pre-exponential factors were evaluated at various stages of the conversion process. Additionally, an optimized artificial neural network (ANN) model was developed to predict thermal degradation at different heating rates. The ANN model successfully predicted thermal degradation data for unseen heating rates with R2 > 0.98. This ANN approach provides significant advantages, reducing the need for extensive experimental work and offering valuable applications for industrial optimization and pyrolysis reactor development.